# Customize /100x for Your Workspace

Paste this entire file into Claude Code. Then paste the contents of `100x.md` directly below it. Send.

---

I'm forking the `/100x` skill from Russ Gardner's Jack's Community case study pack. The skill body is below this prompt. Before installing it, ask me 5 questions and customize the skill so it fits my domain.

## Questions to ask (one at a time, wait for my answer before moving to the next)

1. **Your industry or domain.** "B2B SaaS", "indie game dev", "freelance copywriting", "real estate investing", "healthcare tech", whatever applies. Be specific enough that the lenses can adapt.
2. **Decision archetype you'll most often run this against.** Strategy / marketing / hiring / architecture / pricing / product / positioning / something else. Pick the most common case. Don't worry, the skill still handles other archetypes, this just tunes the default lens set.
3. **Two or three constraints that always apply.** Budget cap ("under $5K total per quarter"), time horizon ("must ship in 30 days"), regulatory ("HIPAA / SOC2"), team size ("solo operator"), tech ("must run on existing Postgres stack"). These get injected into every agent prompt as hard constraints.
4. **Pick 4 to 7 lenses from this list, or write your own:**
   - Technical / Architecture
   - Business / Revenue
   - Competitive / Landscape
   - Growth / Distribution
   - Market / Timing
   - Financial / Unit Economics
   - Operational / Process
   - Regulatory / Compliance
   - Customer / Buyer Psychology
   - Brand / Positioning
   - Risk / Failure Modes
   - User Experience / Design
   - Cultural / Macro Trends
   - (write-in)
   
   The skill spawns one agent per lens. Pick the ones most likely to surface non-obvious insight in your domain.
5. **Where to save outputs.** Pick one:
   - **Supabase / Postgres table** (you'll give the table name and connection details, the skill writes a row per run)
   - **Local folder** (you'll give a path like `~/research/100x/`, the skill saves a numbered Markdown file)
   - **Obsidian vault** (you'll give a folder path inside your vault, the skill writes there with a global running count)
   - **Nowhere** (each run lives in the chat transcript only, acceptable for low-stakes use)

## What to do with my answers

After I've answered all 5, do the following:

1. **Read the skill body below** (everything after the `---` separator at the bottom of this prompt).
2. **Replace every `> FILL IN:` block** with content derived from my answers. Specifically:
   - The "Step 0: Load Research Memory" block: write the actual lookup mechanism for my chosen save location. If Supabase, write the SQL. If folder, write the file-listing logic. If Obsidian, same as folder. If Nowhere, replace the FILL IN block with "Skip this step. This installation does not persist research outputs."
   - The "Step 0.5: Load Business State" block: ask me if I have any permanent rules to inject. If yes, embed them. If I say no or skip, replace the block with "No permanent rules configured. Add them later if patterns emerge."
   - The "Step 1: Decompose the Problem" block: replace the example lens list with my actual chosen lenses from question 4. Keep the bullet format and a one-line description per lens.
   - The "Step 6: Save Output" block: write the actual save path and naming convention. If Obsidian or local folder, hardcode the path. If Supabase, leave the SQL skeleton with my table name filled in. If Nowhere, replace with "This installation does not save outputs to disk. Each run lives in the chat transcript."
   - The "Step 7: Persist" block: same as Step 0, write the actual persistence mechanism. If Nowhere, replace with "This installation does not persist. Skip Step 7."
3. **Embed my constraints** in a permanent block near the top of the file, under a heading like "Hard Constraints". These get auto-injected into every agent prompt. Format them as a bulleted list the agents read as ground truth.
4. **Update the description metadata** in the YAML front matter to mention my industry. So when Claude Code shows this skill in a list, it reads "Spawn parallel research agents for {industry} strategic decisions" or similar.
5. **Save the customized skill** to `~/.claude/skills/100x/SKILL.md`. Create the directory if it doesn't exist. If a file is already there, write the new version to `~/.claude/skills/100x/SKILL.md.new` and tell me to compare before overwriting.
6. **Show me the diff.** Highlight which placeholders got replaced. Confirm I can run `/100x {question}` in any future session.

## Bias toward action

Make sensible defaults if my answers are vague. Don't loop on clarifying questions. The customize step should take 5 minutes.

If I picked "Nowhere" for save location, the skill loses the cross-run compounding benefit. Tell me that once, then proceed. Don't talk me out of it, just make sure I know what I'm trading.

---

(Skill body follows below the next separator. Read it, then run the steps above.)

---